This is mainly a bug fix version of the 3.0 series. All tickets solved are described below.

# Tickets fixed

## Ticket #193: Error in the computation of the consistent tangent operator in the OrthotropicElastic5.mfront and OrthotropicElastic6.mfront example

For more details, see: https://sourceforge.net/p/tfel/tickets/193/

## Ticket #188: Default constructor of the qt class is ill formed: noexcept specifier is not correcty defined

For more details, see: https://sourceforge.net/p/tfel/tickets/188/

## Ticket #185: Problem on transposition of fourth order tensor

The ST2toST2TransposeExpr<A> class expected that class A defines the IndexType typedef, but this is not a requirement of the ST2toST2Concept concept. Thus the transpose operation did not work for all classes matching this concept. In particular, the following code used to fail:

const auto tM2 = transpose(Stensor4::Id() * Stensor4::Id());

For more details, see: https://sourceforge.net/p/tfel/tickets/184/

## Ticket #181: Runge-Kutta DSL does not handle NaN values

For most algorithms with correction-prediction, the Runge-Kutta DSL has two strategies to evaluate the error. The first one (the default) is the euclidian norm and the second one the based on the absolute norm. The DSL switches from one strategy to the other depending on the size of unknowns.

The trouble only concerns the second strategy.

For more details, see: https://sourceforge.net/p/tfel/tickets/181/

## Ticket #180: “return false” doesn’t work with some Runge-Kutta integrators

The issue concerns all the rk algorithms without local time-step control (euler, rk2, rk4) which are not used in practice.

For more details, see: https://sourceforge.net/p/tfel/tickets/180/

## Ticket #178: Name of the python module is wrong under Mac OS

For more details, see: https://sourceforge.net/p/tfel/tickets/178/

## Ticket #175: Add rpathto the TFEL libraries when generating the library on Mac Os

For more details, see: https://sourceforge.net/p/tfel/tickets/175/

## Ticket #170: Default parameters’ values precision

The default parameters’ values were truncated to 6 significant digits (default number of digits of the std::ostream class).

For more details, see: https://sourceforge.net/p/tfel/tickets/170/

# Ticket #133: Equation giving dn_ddai page 20 in the tutorial is wrong

For more details, see: https://sourceforge.net/p/tfel/tickets/133/